Transcription initiation platforms and GTF recruitment at tissue-specific enhancers and promoters

Journal name:
Nature Structural & Molecular Biology
Year published:
Published online


Recent work has shown that RNA polymerase (Pol) II can be recruited to and transcribe distal regulatory regions. Here we analyzed transcription initiation and elongation through genome-wide localization of Pol II, general transcription factors (GTFs) and active chromatin in developing T cells. We show that Pol II and GTFs are recruited to known T cell–specific enhancers. We extend this observation to many new putative enhancers, a majority of which can be transcribed with or without polyadenylation. Importantly, we also identify genomic features called transcriptional initiation platforms (TIPs) that are characterized by large areas of Pol II and GTF recruitment at promoters, intergenic and intragenic regions. TIPs show variable widths (0.4–10 kb) and correlate with high CpG content and increased tissue specificity at promoters. Finally, we also report differential recruitment of TFIID and other GTFs at promoters and enhancers. Overall, we propose that TIPs represent important new regulatory hallmarks of the genome.

At a glance


  1. Pol II and GTF recruitment to T-cell stage-specific enhancers of active loci or genes poised for activation.
    Figure 1: Pol II and GTF recruitment to T-cell stage-specific enhancers of active loci or genes poised for activation.

    (ac) ChIP-seq binding profiles for GTFs, CBP, total, initiating (Ser5P) and elongating (Ser2P) Pol II, active chromatin marks and FAIRE (accessible DNA regions). Light gray vertical bands show previously annotated or characterized enhancer regions, and dark gray bands indicate promoters. Normalized ChIP-seq signals for each experiment are shown on the right. Conservation, mappability track, regulatory elements and genes on positive (+) or negative (−) strand are indicated below the ChIP-seq lanes. In a and b at the active Cd4 and Cd8 loci, GTFs and initiating (Ser5P) Pol II are detected at proximal enhancer (PE) and thymocyte enhancer (TE), as opposed to the distal enhancer (DE), elements (Cd4) and EI, EII and EV (Cd8). At the Il2ra inactive locus (c), poised for transcription and activated either before or after the double-positive differentiation stage, Ser5P and GTFs are detected at a previously characterized enhancer (element V) region. The binding profiles of the remaining GTFs are shown in Supplementary Figure 2a–c.

  2. Epigenetic or transcriptional features and tissue-specific expression of putative enhancers recruiting Ser5P and TBP.
    Figure 2: Epigenetic or transcriptional features and tissue-specific expression of putative enhancers recruiting Ser5P and TBP.

    (a) Average binding profiles of GTFs, Pol II, active chromatin marks, CBP and FAIRE in IGRs (centered on the maximum TBP signal) and non-oriented promoters (centered on the TSS; see Supplementary Fig. 3d for oriented genes). (b) Tissue-specific expression of genes associated with promoters or adjacent to putative enhancer IGRs. Using microarray data, associated genes were analyzed for their expression in various tissues and ordered by decreasing ratio with the whole genome (tissues with the highest five ratios are shown; see Supplementary Fig. 4 for all tissues). (c) Genes associated with TBP and Ser5P IGRs show a more tissue-restricted expression compared to promoters and to other selections. Box plot of expression for IGRs and promoters in double-positive cells and the remaining tissues are shown. The differential (red and blue bars) is greater for IGRs (left). The same analysis was conducted using different IGR-selection criteria including H3K4me1 and H3K4me3 (yellow), CBP and H3K4me1 (orange) and CBP-H3K4me1-H3K4me3 (green). Our TBP-Ser5P-H3K4me1-H3K4me3 (blue) selection showed the highest tissue-restricted expression as well as the highest expression levels of IGR-associated genes (right). Similar analyses using all hematopoietic tissues are shown in Supplementary Figure 4c. (d) Validation of enhancer activity of two TBP and Ser5P IGRs in a promoter-dependent luciferase reporter assay. The Dusp6 and Rhoh promoters and IGRs were cloned in a pGL3 vector and transfected in a T-cell line (EL4). The Dusp6 IGR was cloned into both orientations and retained its ability to enhance promoter-driven expression. Error bars represent s.e.m. from two independent transfections. The complete experiment is presented in Supplementary Figure 11c.

  3. TBP and Ser5P enhancers are transcribed with or without polyadenylation.
    Figure 3: TBP and Ser5P enhancers are transcribed with or without polyadenylation.

    (a) Examples of known enhancers transcribed in the presence (E8I element of the Cd8 locus, left) or absence (PE element of the Cd4 locus, right) of polyadenylation signal. Total and poly(A) RNAs signals are represented below the TBP and Ser5P ChIP-seq lanes as log2 signals of the directional RNA-seq experiments. RNA strand orientation is indicated on the left. Transcribed enhancer elements are indicated by a light gray vertical bands. Additional examples of IGR transcription are shown in Supplementary Figure 7. Possible tracking of Pol II toward Cd8a is indicated by the dotted arrow below the Ser5P lane. (b) Pol II ChIP-seq and oriented RNA-seq average profiling on TBP and Ser5P enhancer IGRs or gene promoters (left panels) for total (top panels) or poly(A) RNAs (bottom panels). Selected TBP and Ser5P IGRs were divided into three populations associated with either poly(A), no poly(A) or no RNA, and orientation of the IGRs was established based on the RNA levels. Signals are centered on the TSS of genes or on the main TBP peak of IGRs.

  4. Poly(A) and non-poly(A) IGR subpopulations show distinct chromatin signatures between each other and genes.
    Figure 4: Poly(A) and non-poly(A) IGR subpopulations show distinct chromatin signatures between each other and genes.

    Comparison of active chromatin marks, CBP and ETS1 on oriented IGRs and promoters. ChIP-seq average binding profiles of the IGR populations described in Figure 3b are shown for genes and poly(A) (upper panels), no poly(A) and no RNA IGRs (lower panels). The profiles for the remaining factors described in this study are shown in Supplementary Figure 7d.

  5. Pol II and GTFs transcription initiation platforms.
    Figure 5: Pol II and GTFs transcription initiation platforms.

    (a) Examples of transcriptional initiation (TBP, Ser5P) or elongation (Ser2P, H3K36me3) hallmarks on TIPs at promoters, intragenic or IGR locations from left to right, respectively. The TBP and Ser5P TIPs isolated using a systematic approach (see Online Methods) are indicated by a red horizontal bar below TBP and Ser5P ChIP-seq signals. (b) Heatmaps of TIPs sorted by size and anchored on their center. TIPs (based on TBP and Ser5P selection) are shown for TBP, Ser5P and for the corresponding profiles for GTFs, active chromatin marks, FAIRE and the RNA-seq signal for positive and negative strands. TIP boundaries are represented by a red 5′ (right side) and green (left side) 3′ line. Heatmaps for all ChIP-seq and RNA-seq experiments at promoters, intragenic and IGR locations (as well as for input or mock Ig controls) are included in Supplementary Figure 9b,c.

  6. TIPs correlate with CpG content and tissue-specific expression at promoters.
    Figure 6: TIPs correlate with CpG content and tissue-specific expression at promoters.

    (a) CpG content across TIPs anchored on their center, similarly to Figure 5b. Promoters clearly show the highest CpG content. Similar trends are also visible in IGRs and, to a lesser extent, in intragenic regions. (b) Clustering of T-cell and non–T-cell transcription factor motifs at TIPs around promoters. Most putative TFBS overlap with the high CpG content from a. (c) Analysis of tissue specificity of expression for genes associated to promoter TIPs, similarly to Figure 2b. The associated genes show a more pronounced double-positive T-cell gene-expression pattern, indicating an increased tissue specificity at TIPs. The remaining genomic regions are analyzed in Supplementary Figure 10a. (d) Genes associated with promoter TIPs were classed into four equally sized groups (quartiles) with increasing platform size. The tissue specificity of the expression pattern increases with platform size, as indicated by the increasing rank and decreasing associated P values. The complete ranks are presented in Supplementary Figure 10b. (e) Correlation of TIP size to absolute gene expression levels at promoters (r = 0.33). The global and local fitted curves are represented by solid red and dashed black lines, repectively. Similar graphs for IGRs and intragenic regions are shown in Supplementary Figure 10c. TIP size and expression values were transformed using a hyperbolic arcsine (Asinh) function.

  7. Average profiles of TIPs and model summarizing their features at distinct genomic locations.
    Figure 7: Average profiles of TIPs and model summarizing their features at distinct genomic locations.

    (a) Average profiles of TBP, Ser5P, TFIIB, TFIIH, TFIIE, active chromatin marks, ETS1 and FAIRE across all resized TIPs. Regions were divided into promoter (red), IGR (blue) and intragenic (green) locations (see Supplementary Fig. 12a for total profiles). In general, Ser5P, GTFs, ETS1, FAIRE and H3K4me3 are largely enriched throughout the platforms, H3K4me1 peaks just after the boundaries and H3K36me3 is depleted. IGR and intragenic profiles show mostly similar patterns, including lower TBP, Ser5P and H3K4me3 as well as higher TFIIH, TFIIE, H3K4me1 and ETS1 levels, compared to promoters. The remaining profiles are shown in Supplementary Figure 12b. (b) Total RNA signal across the different classes of TIPs. RNAs from the positive (blue) and negative (red) strands are shown. Transcription appears to start at either border, increases toward the opposite boundary and decreases again afterwards, indicative of a transcriptional barrier possibly imposed by H3K4me1. (c) Promoter, IGRs and intragenic TIPs are all characterized by open chromatin regions and are delimited by an enrichment of the H3K4me1 (Me1, green circles) histone mark (to a lesser extent at promoter), possibly reflecting a nucleosomal barrier. These areas have transcription initiation hallmarks such as Ser5P Pol II (green), H3K4me3 (Me3, yellow circles)—though less pronounced in IGRs and intragenic regions—and GTF recruitment in common. They differ in their relative proportions of ETS1 (designated here as TF) and GTFs at IGRs, compared to promoters. Promoters fundamentally differ from other TIPs in their ability to allow Pol II to enter elongation (blue), although Ser2P is not detected in the immediate proximity of TIPs (most likely because of higher elongation rate and less Ser2P accumulation at the 5′ ends). CpG and TFBS are more prominent at promoters, although TFs, such as ETS1, are more often recruited to enhancers. Bidirectional transcription is present at both promoter37, 38 and IGRs (Fig. 3b and Supplementary Fig. 9d).

Accession codes

Referenced accessions

Gene Expression Omnibus


  1. Sikorski, T.W. & Buratowski, S. The basal initiation machinery: beyond the general transcription factors. Curr. Opin. Cell Biol. 21, 344351 (2009).
  2. Buratowski, S. Progression through the RNA polymerase II CTD cycle. Mol. Cell 36, 541546 (2009).
  3. Carninci, P. et al. Genome-wide analysis of mammalian promoter architecture and evolution. Nat. Genet. 38, 626635 (2006).
  4. Koch, F., Jourquin, F., Ferrier, P. & Andrau, J.C. Genome-wide RNA polymerase II: not genes only!. Trends Biochem. Sci. 33, 265273 (2008).
  5. Heintzman, N.D. et al. Histone modifications at human enhancers reflect global cell-type-specific gene expression. Nature 459, 108112 (2009).
  6. Heintzman, N.D. et al. Distinct and predictive chromatin signatures of transcriptional promoters and enhancers in the human genome. Nat. Genet. 39, 311318 (2007).
  7. Wang, Z. et al. Combinatorial patterns of histone acetylations and methylations in the human genome. Nat. Genet. 40, 897903 (2008).
  8. Barski, A. et al. High-resolution profiling of histone methylations in the human genome. Cell 129, 823837 (2007).
  9. Visel, A. et al. ChIP-seq accurately predicts tissue-specific activity of enhancers. Nature 457, 854858 (2009).
  10. Xi, H. et al. Identification and characterization of cell type-specific and ubiquitous chromatin regulatory structures in the human genome. PLoS Genet. 3, e136 (2007).
  11. Higgs, D.R., Vernimmen, D. & Wood, B. Long-range regulation of alpha-globin gene expression. Adv. Genet. 61, 143173 (2008).
  12. Fromm, G. & Bulger, M. A spectrum of gene regulatory phenomena at mammalian beta-globin gene loci. Biochem. Cell Biol. 87, 781790 (2009).
  13. Szutorisz, H., Dillon, N. & Tora, L. The role of enhancers as centres for general transcription factor recruitment. Trends Biochem. Sci. 30, 593599 (2005).
  14. De Santa, F. et al. A large fraction of extragenic RNA pol II transcription sites overlap enhancers. PLoS Biol. 8, e1000384 (2010).
  15. Kim, T.K. et al. Widespread transcription at neuronal activity-regulated enhancers. Nature 465, 182187 (2010).
  16. Giresi, P.G., Kim, J., McDaniell, R.M., Iyer, V.R. & Lieb, J.D. FAIRE (Formaldehyde-Assisted Isolation of Regulatory Elements) isolates active regulatory elements from human chromatin. Genome Res. 17, 877885 (2007).
  17. Bosselut, R. CD4/CD8-lineage differentiation in the thymus: from nuclear effectors to membrane signals. Nat. Rev. Immunol. 4, 529540 (2004).
  18. Anderson, M.K. At the crossroads: diverse roles of early thymocyte transcriptional regulators. Immunol. Rev. 209, 191211 (2006).
  19. Sawada, S. & Littman, D.R. Identification and characterization of a T-cell-specific enhancer adjacent to the murine CD4 gene. Mol. Cell Biol. 11, 55065515 (1991).
  20. Cherrier, M., D'Andon, M.F., Rougeon, F. & Doyen, N. Identification of a new cis-regulatory element of the terminal deoxynucleotidyl transferase gene in the 5′ region of the murine locus. Mol. Immunol. 45, 10091017 (2008).
  21. Kaufmann, C. et al. A complex network of regulatory elements in Ikaros and their activity during hemo-lymphopoiesis. EMBO J. 22, 22112223 (2003).
  22. Georgopoulos, K., van den Elsen, P., Bier, E., Maxam, A. & Terhorst, C.A. T cell-specific enhancer is located in a DNase I-hypersensitive area at the 3′ end of the CD3-delta gene. EMBO J. 7, 24012407 (1988).
  23. Yannoutsos, N. et al. A cis element in the recombination activating gene locus regulates gene expression by counteracting a distant silencer. Nat. Immunol. 5, 443450 (2004).
  24. Hostert, A. et al. Hierarchical interactions of control elements determine CD8alpha gene expression in subsets of thymocytes and peripheral T cells. Immunity 9, 497508 (1998).
  25. Ellmeier, W., Sunshine, M.J., Losos, K. & Littman, D.R. Multiple developmental stage-specific enhancers regulate CD8 expression in developing thymocytes and in thymus-independent T cells. Immunity 9, 485496 (1998).
  26. Schmidl, C. et al. Lineage-specific DNA methylation in T cells correlates with histone methylation and enhancer activity. Genome Res. 19, 11651174 (2009).
  27. Schorle, H., Holtschke, T., Hunig, T., Schimpl, A. & Horak, I. Development and function of T cells in mice rendered interleukin-2 deficient by gene targeting. Nature 352, 621624 (1991).
  28. Shi, W. & Zhou, W. Frequency distribution of TATA Box and extension sequences on human promoters. BMC Bioinformatics 7 (suppl. 4), S2 (2006).
  29. Eyquem, S., Chemin, K., Fasseu, M. & Bories, J.C. The Ets-1 transcription factor is required for complete pre-T cell receptor function and allelic exclusion at the T cell receptor beta locus. Proc. Natl. Acad. Sci. USA 101, 1571215717 (2004).
  30. Xie, X. et al. Systematic discovery of regulatory motifs in human promoters and 3′ UTRs by comparison of several mammals. Nature 434, 338345 (2005).
  31. Saxonov, S., Berg, P. & Brutlag, D.L. A genome-wide analysis of CpG dinucleotides in the human genome distinguishes two distinct classes of promoters. Proc. Natl. Acad. Sci. USA 103, 14121417 (2006).
  32. Blackledge, N.P. et al. CpG islands recruit a histone H3 lysine 36 demethylase. Mol. Cell 38, 179190 (2010).
  33. Spicuglia, S. et al. Promoter activation by enhancer-dependent and -independent loading of activator and coactivator complexes. Mol. Cell 10, 14791487 (2002).
  34. Ho, Y., Elefant, F., Cooke, N. & Liebhaber, S. A defined locus control region determinant links chromatin domain acetylation with long-range gene activation. Mol. Cell 9, 291302 (2002).
  35. Andrau, J.C. et al. Genome-wide location of the coactivator mediator: Binding without activation and transient Cdk8 interaction on DNA. Mol. Cell 22, 179192 (2006).
  36. Radonjic, M. et al. Genome-wide analyses reveal RNA polymerase II located upstream of genes poised for rapid response upon S. cerevisiae stationary phase exit. Mol. Cell 18, 171183 (2005).
  37. Core, L.J., Waterfall, J.J. & Lis, J.T. Nascent RNA sequencing reveals widespread pausing and divergent initiation at human promoters. Science 322, 18451848 (2008).
  38. Seila, A.C. et al. Divergent transcription from active promoters. Science 322, 18491851 (2008).
  39. Ørom, U.A. et al. Long noncoding RNAs with enhancer-like function in human cells. Cell 143, 4658 (2010).
  40. Boyer, L.A. et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 122, 947956 (2005).
  41. Nicol, J.W., Helt, G.A., Blanchard, S.G. Jr., Raja, A. & Loraine, A.E. The Integrated Genome Browser: free software for distribution and exploration of genome-scale datasets. Bioinformatics 25, 27302731 (2009).
  42. Benoukraf, T. et al. CoCAS: a ChIP-on-chip analysis suite. Bioinformatics 25, 954955 (2009).
  43. Wu, C. et al. BioGPS: an extensible and customizable portal for querying and organizing gene annotation resources. Genome Biol. 10, R130 (2009).
  44. Bailey, T.L. & Elkan, C. Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc. Int. Conf. Intell. Syst. Mol. Biol. 2, 2836 (1994).
  45. Sandelin, A., Alkema, W., Engstrom, P., Wasserman, W.W. & Lenhard, B. JASPAR: an open-access database for eukaryotic transcription factor binding profiles. Nucleic Acids Res. 32, D91D94 (2004).
  46. Thomas-Chollier, M. et al. RSAT: regulatory sequence analysis tools. Nucleic Acids Res. 36, W119W127 (2008).

Download references

Author information

  1. These authors contributed equally to this work.

    • Frederic Koch,
    • Romain Fenouil &
    • Marta Gut


  1. Centre d'Immunologie de Marseille-Luminy, Université Aix-Marseille, Campus de Luminy, Marseille, France.

    • Frederic Koch,
    • Romain Fenouil,
    • Pierre Cauchy,
    • Joaquin Zacarias-Cabeza,
    • Salvatore Spicuglia,
    • Albane Lamy de la Chapelle,
    • Pierre Ferrier &
    • Jean-Christophe Andrau
  2. Centre National de la Recherche Scientifique, UMR6102, Marseille, France.

    • Frederic Koch,
    • Romain Fenouil,
    • Pierre Cauchy,
    • Joaquin Zacarias-Cabeza,
    • Salvatore Spicuglia,
    • Albane Lamy de la Chapelle,
    • Pierre Ferrier &
    • Jean-Christophe Andrau
  3. Institut National de la Santé et de la Recherche Médicale, U631, Marseille, France.

    • Frederic Koch,
    • Romain Fenouil,
    • Pierre Cauchy,
    • Joaquin Zacarias-Cabeza,
    • Salvatore Spicuglia,
    • Albane Lamy de la Chapelle,
    • Pierre Ferrier &
    • Jean-Christophe Andrau
  4. Centre National de Génotypage, Commissariat à l'Energie Atomique, Evry, France.

    • Marta Gut &
    • Ivo Gut
  5. Fondation Jean Dausset—Centre d'Etude du Polymorphisme Humain, Paris, France.

    • Marta Gut
  6. Centre Nacional D'Anàlisi Genòmica, Parc Científic de Barcelona, Baldiri i Reixac, Barcelona, Spain.

    • Marta Gut &
    • Ivo Gut
  7. Techniques Avancées pour le Génome et la Clinique, Marseille, France.

    • Pierre Cauchy
  8. Institute of Molecular Tumor Biology, Medical Faculty of the Westfälische Wilhelms-Universität, Münster, Germany.

    • Thomas K Albert
  9. Department of Molecular Epigenetics, Helmholtz Center Munich, Center of Integrated Protein Science, Munich, Germany.

    • Martin Heidemann,
    • Corinna Hintermair &
    • Dirk Eick


J.-C.A., F.K., T.K.A., P.F. and I.G. conceived the framework of the study. J.-C.A. and F.K. designed the experiments. R.F., P.C. and F.K. carried out the bioinformatic analyses and data treatment. D.E., C.H. and M.H. produced and provided the Ser2P and Ser5P antibodies as well as other antibodies that were not presented in this study. All ChIP-seq and RNA-seq materials were prepared by F.K. with the exception of ETS1 ChIP-seq, which was prepared by P.C., M.G. and I.G. conducted all ChIP-seq and RNA sequencing experiments. J.Z.-C. and S.S. did the FAIRE experiment. F.K. did the cloning and luciferase experiments and A.L.d.l.C. participated and provided technical assistance. J.-C.A. wrote the manuscript, and F.K., R.F. and P.C. participated in its preparation. All authors reviewed the manuscript.

Competing financial interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to:

Author details

Supplementary information

PDF files

  1. Supplementary Text and Figures (10M)

    Supplementary Figures 1–12, Supplementary Tables 1 and 2, and Supplementary Methods

Additional data